Efficient retrieval of the top-k most relevant spatial web objects

Abstract

The conventional Internet is acquiring a geo-spatial dimension. Web documents are being geo-tagged, and geo-referenced objects such as points of interest are being associated with descriptive text documents. The resulting fusion of geo-location and documents enables a new kind of top-k query that takes into account both location proximity and text relevancy. To our knowledge, only naive techniques exist that are capable of computing a general web information retrieval query while also taking location into account.This paper proposes a new indexing framework for locationaware top-k text retrieval. The framework leverages the inverted file for text retrieval and the R-tree for spatial proximity querying. Several indexing approaches are explored within the framework. The framework encompasses algorithms that utilize the proposed indexes for computing the top-k query, thus taking into account both text relevancy and location proximity to prune the search space. Results of empirical studies with an implementation of the framework demonstrate that the paper's proposal offers scalability and is capable of excellent performance.

On Generalizing Collective Spatial Keyword Queries

“…A top-k kNN query [8], [18], [15], [19], [20], [9], [25] adopts the ranking function considering both the spatial proximity and the textual relevance of the objects and returns top-k objects based on the ranking function. This type of queries has been studied on Euclidean space [8], [18], [15], road network databases [19], trajectory databases [20], [9] and moving object databases [25].…”

“…This type of queries has been studied on Euclidean space [8], [18], [15], road network databases [19], trajectory databases [20], [9] and moving object databases [25]. Usually, the methods for this kind of queries adopt an index structure called the IR-tree [8], [23] capturing both the spatial proximity and the textual information of the objects to speed up the keyword-based nearest neighbor (NN) queries and range queries. In this paper, we also adopt the IR-tree for keyword-based NN queries and range queries.…”

“…The procedure is presented in in Algorithm 4, and it is similar to the "findBestFeasibleSet" procedure (in Algorithm 2) except that it replaces the enumeration process with an iterative process (lines [8][9][10][11][12][13][14] for searching for a feasible set.…”

Location-Based Instant Search

“…There are recent studies on integrating a spatial index with a keyword index [11,9,8,20,21,19,14]. The proposed methods add a keyword filter to each node in the spatial tree node that describes the keyword information in the subtree of that node.…”

“…The work in [11] constructs a global inverted index to map from keywords to tree nodes that have these keywords. These studies consider the problem of range search, nearest neighbor search [12,17], or top-k search [8]. The work in [8] proposed two ideas to improve the performance of searching with this type of indices: using object similarities to influence the structure of the tree index, and creating clusters for similar objects and indexing on them.…”